Of course one thing not brought about social welfare is that isn't a choice of to pay or not but to pay now vs pay later no matter what you do - in addition to humanitarian and reputation impacts.
It is harder to calculate given hard numbers only exist if things go wrong.
Not paying for vaccination or hygenic programs leads to outbreaks which are expensive to contain and not containing leads to even more.
Even literally leaving people to die in the street still generates expenses
The evidence shows that social welfare spending can increase future costs. For example it increases the percentage of women who have children out of wedlock. The increase in social welfare spending is the primary reason why the percentage of children born to single parents rose from about 4% in 1948 to 40% today. That in turn increases nearly every type of socioeconomic problem.
Maybe people just don't put as much importance on weddings as they did when children out of wedlock had substantially reduced rights? I don't think the link to increased welfare spending is as clear as you make it. I for example only married for the tax benefits.
The evidence suggests that increases in the availability of social assistance cause an increase in the percentage of children born out of wedlock, so it's not merely an observation of a correlation.
There are undoubtedly other factors as well, but the expansion of social spending is one of the major ones.
>Public assistance also had large, statistically significant effects on all of the dependent variables. All else equal, higher levels of public assistance were associated with lower prevalence of marriage for black men and black women, lower prevalence of husband-wife families, lower percentage of marital births for black women, and lower percentage of black children living in husband-wife families. These results differed from those reported in some previous aggregate-level studies (e.g., Ellwood & Bane, 1985; Ellwood & Summers, 1986), but they were quite robust and were substantively as well as statistically significant. For example, the difference between $150 and $250 in average support per recipient child (a difference of 0.51 in the natural log version, which is only slightly larger than the sample standard deviation for the variable) translates into a 4.4-point difference in the expected percentage of marital births for black women aged 20-24, a difference of nearly 2.4 points in the expected percentage of black children residing in husband-wife families, and a 3.5 point difference in the expected percentage of married for women with children 0-5.
>These estimated impacts were then combined to measure "final" effects on the total illegitimacy rate - out-of-wedlock births per 1,000 women in the population. The outcome of this experiment was a hypothetical decrease in total illegitimacy rates - as of the end of the estimation period in 1992 - of about 15 percent for whites and slightly under 7 percent for blacks. The differential effects are almost entirely accounted for by the disparities in the determinants of marital status noted above.
>
>Narrowly defined, our goal in this study has been to test the linkage between welfare benefits and illegitimate births. The econometric evidence presented here has strongly supported this connection. More broadly, however, this study implicates the welfare system in the growth of illegitimacy
It is harder to calculate given hard numbers only exist if things go wrong.
Not paying for vaccination or hygenic programs leads to outbreaks which are expensive to contain and not containing leads to even more. Even literally leaving people to die in the street still generates expenses